U.S. patent number 5,286,968 [Application Number 07/834,226] was granted by the patent office on 1994-02-15 for method and device for multichannel analog detection.
This patent grant is currently assigned to Centre National de la Recherche Scientifique (CNRS). Invention is credited to Albert-Claude Boccara, Francois Charbonnier, Daniele Fournier, Philippe Gleyzes.
United States Patent |
5,286,968 |
Fournier , et al. |
February 15, 1994 |
Method and device for multichannel analog detection
Abstract
The invention relates to a device for multichannel analog
detection of a signal to be detected having a very good
signal/noise ratio. It incorporates a modulator (53) producing a
modulated signal S(P); means of synchronous attenuation (54) of
variable phase .PHI. producing an attenuated modulated signal; a
multipoint receiver (52) receiving the modulated-attenuated signal
and producing for each point a primary analog signal; an integrator
producing for each point a value V(P,.PHI.) resulting from the
integration over N periods of the primary analog signal; means of
reading, of digitizing and of storing the values V(P,.PHI.) for a
given .PHI. value; a phase sequencer giving .PHI. the values
.PHI..sub.0 +i2.pi./n successively where i is an integer varying
from 1 to n; a digital processing unit making it possible to obtain
data representative of S(P) from the values V(P,.PHI.). It is
particularly well adapted to the detection of a luminous flux with
an array of photodiodes.
Inventors: |
Fournier; Daniele (Paris,
FR), Charbonnier; Francois (Paris, FR),
Gleyzes; Philippe (Paris, FR), Boccara;
Albert-Claude (Paris, FR) |
Assignee: |
Centre National de la Recherche
Scientifique (CNRS) (Paris, FR)
|
Family
ID: |
9398175 |
Appl.
No.: |
07/834,226 |
Filed: |
March 10, 1992 |
PCT
Filed: |
June 27, 1991 |
PCT No.: |
PCT/FR91/00514 |
371
Date: |
March 10, 1992 |
102(e)
Date: |
March 10, 1992 |
PCT
Pub. No.: |
WO92/00549 |
PCT
Pub. Date: |
January 09, 1992 |
Foreign Application Priority Data
|
|
|
|
|
Jun 29, 1990 [FR] |
|
|
90 08255 |
|
Current U.S.
Class: |
250/208.1;
250/214R |
Current CPC
Class: |
G01J
9/02 (20130101); G01B 11/303 (20130101) |
Current International
Class: |
G01B
11/30 (20060101); G01J 9/02 (20060101); G01J
9/00 (20060101); H01J 040/14 () |
Field of
Search: |
;250/214R,208.1,216,234
;356/359,351 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
Other References
"Quantitative Surface topography determination by Nomarski
reflection microscopy. I. Theory", Lessor et al., Opt Soc. America,
vol. 69, No. 2. pp. 357-366..
|
Primary Examiner: Nelms; David C.
Assistant Examiner: Allen; S. B.
Attorney, Agent or Firm: Jacobson, Price, Holman &
Stern
Claims
What is claimed is:
1. Method for multichannel analog detection of a signal to be
detected in which:
the signal to be detected is modulated, producing a modulated
signal S(P) of frequency (f) accompanied by noise;
the modulated signal is subjected to a synchronous attenuation of
phase .PHI. producing an attenuated modulated signal;
the modulated-attenuated signal is received by a multipoint
receiver which produces, for each point, a primary analog
electrical signal;
each primary analog signal is integrated over N periods producing,
for each point (P), a value V(P,.PHI.);
the values V(P,.PHI.), for the set of points and for a fixed phase
value .PHI., are read, digitized, stored;
the phase .PHI. is successively increased (n-1) times by 2.pi./n,
and for each of these values the operations which make it possible
to store the values V(P,.PHI.) are repeated;
the n values V(P,.PHI.) obtained respectively for each point (P)
are digitally processed in such a way as to produce a data item
representative of the amplitude and of the phase of S(P).
2. Method of multichannel analog detection according to claim 1,
characterized in that the phase .PHI. is increased successively by
.pi./2(n=4).
3. Method of multichannel analog detection according to claim 1,
characterized in that the synchronous attenuation is produced by
the multiplication of the modulated signal by a square-wave
function successively taking the values 0 and 1.
4. Method of multichannel analog detection according to claim 1,
characterized in that the signal to be detected is a luminous
signal.
5. Method of multichannel analog detection according to claim 4,
characterized in that the modulation of the signal to be detected
is obtained by modulation of the luminous source.
6. Method of multichannel analog detection according to claim 1,
characterized in that the primary analog electrical signal, for
each point, is obtained by heterodyne detection.
7. Device for multichannel analog detection of a signal to be
detected comprising:
a modulator (53) producing a modulated signal S(P);
means of synchronous attenuation (54) of variable phase .PHI.
producing an attenuated modulated signal;
a multipoint receiver (52) receiving the modulated-attenuated
signal and producing for each point a primary analog signal;
an integrator producing for each point a value V(P,.PHI.) resulting
from the integration over N periods of the primary analog
signal;
means of reading, of digitizing and of storing the values
V(P,.PHI.) for a given .PHI. value;
a phase sequencer giving .PHI. the values .PHI..sub.0 +i2.pi./n
successively where i is an integer varying from 1 to n;
a digital processing unit making it possible to obtain data
representative of S(P) from the values V(P,.PHI.).
8. Device for multichannel analog detection according to claim 7,
characterized in that:
the signal to be detected is a luminous signal carried by a beam
produced by a luminous source;
it incorporates photodiodes (CCD) constituting the multipoint
receiver and the integrator.
9. Device for multichannel analog detection according to claim 8,
characterized in that the luminous source is a laser diode
incorporating a supply and in that the means of synchronous
attenuation acts on the supply of the source.
10. Device for multichannel analog detection according to claim 8,
characterized in that the means of synchronous attenuation are a
shutter placed between the luminous source and the multipoint
receiver.
11. Device for multichannel analog detection according to claim 7,
characterized in that it incorporates means of heterodyne detection
making it possible to improve the signal/noise ratio of the primary
analog electrical signal produced for each point by the multipoint
receiver.
12. Device for multichannel analog detection according to claim 7,
characterized in that the means of reading of the values V(P,.PHI.)
produce a signal of the video type and address it to the means of
digitizing.
13. Device for multichannel analog detection according to claim 7,
characterized in that it is used to carry out measurements of
difference in path length by interferential microscopy in polarized
light.
Description
The present invention relates to a method and a device for
multichannel analog detection.
Multichannel detectors such as arrays or matrices of photodiodes
are developing very rapidly and are now in current use.
The methods of detection frequently used with single-channel
detectors in order to improve their signal/noise ratio are often
difficult to transpose directly to multichannel detectors. For this
reason the development of a method and of a device for multichannel
analog detection assumes great significance.
In particular, the method of synchronous detection is currently
used with single-channel detectors. It gives complete satisfaction
and its significance needs no further justification.
According to this method, which can be used equally well for
spectrometers, dichrometers, polarimeters, ellipsometers, optical
or scanning electron microscopes, interferential microscopes, etc.,
a source, most often a luminous source, is modulated. After
detection, the modulated signal is amplified then multiplied by a
reference whose mean value is zero. A low-pass filter followed by a
DC amplifier makes it possible to obtain the information with a
good noise rejection. A rejection rate of more than 90 decibels is
thus frequently obtained.
Having regard to the significance of this method, an attempt has
been made to transpose it for use in multichannel detectors.
To this end, a first solution consists in coupling as many
synchronous detectors as there are elements in the multichannel
detector. This solution requires the parallel reading of the
specialized circuit and rapidly becomes impossible to employ for
detectors comprising a large number of elements.
As an indication, multichannel detectors, which are currently
widespread, are arrays or matrices of diodes incorporating from 256
to several million channels.
A second solution is to read each of the elements of the
multichannel detector in series at a frequency which is higher than
the modulation frequency, to digitize the information, to store it
in various memories, and to process the signal thus derived using a
programmed ADP device.
This solution is limited at present by the processing time of the
digital memories and can only be employed with very low frequencies
of modulation of the signal (a few Hertz). Moreover, the mean being
found digitally, it is necessary to use analog-digital converters
of very high resolution (greater than or equal to 12 bits) in order
to obtain a good dynamic range and a reasonable speed of
measurement. Such converters are very expensive.
Similar difficulties are encountered in the employment of
heterodyne detections applied to the multichannel detectors.
The objective of the invention is thus to propose a method and a
device for multichannel analog detection which are simple to
employ, which improve the signal/noise ratio of the measured signal
and exhibit a good dynamic range.
To this end, there is proposed a method for multichannel analog
detection of a signal to be detected in which:
the signal to be detected is modulated, producing a modulated
signal S(P) of frequency f accompanied by noise;
the modulated signal is subjected to a synchronous attenuation of
phase .PHI. producing an attenuated modulated signal;
the modulated-attenuated signal is received by a multipoint
receiver which produces, for each point, a primary analog
electrical signal;
each primary analog signal is integrated over N periods producing,
for each point P, a value V(P,.PHI.);
the values V(P,.PHI.), for the set of points and for a fixed phase
value .PHI., are read, digitized, stored;
the phase .PHI. is successively increased (n-1) time by 2.pi./n,
and for each of these values the operations which make it possible
to store the values V(P,.PHI.) are repeated;
the n values V(P,.PHI.) obtained respectively for each point P are
digitally processed in such a way as to produce a data item
representative of the amplitude and of the phase of S(P).
The invention also relates to a device for multichannel analog
detection of a signal to be detected comprising:
a modulator producing a modulated signal S(P);
means of synchronous attenuation of variable phase .PHI. producing
an attenuated modulated signal;
a multipoint receiver receiving the modulated-attenuated signal and
producing for each point a primary analog signal;
an integrator producing for each point a value V(P,.PHI.) resulting
from the integration over N periods of the primary analog
signal;
means of reading, of digitizing and of storing the values
V(P,.PHI.) for a given .PHI. value;
a phase sequencer giving .PHI. the values .PHI..sub.0 +i2.pi./n
successively where i is an integer varying from 1 to n;
a digital processing unit making it possible to obtain data
representative of S(P) from the values V(P,.PHI.).
The invention will be described in more detail with reference to
the attached Figures in which:
FIG. 1 is a schematic representation of a multipoint receiver used
according to the invention.
FIGS. 2A-2E are time diagrams of the method of detection according
to the invention.
FIG. 3 is an overall representation of the detection device
according to the invention.
FIG. 4 represents the detection device of the invention in a first
example of an application.
FIG. 5 represents a second example of an application of the
detection device according to the invention.
FIG. 6 is a detailed schematic view of the detection device of the
invention.
FIGS. 7A, 7B and 7C are a diagram of a device for microscopy in
polarized light with which the detection device of the invention is
used.
FIG. 8 is an example of an image obtained with devices of FIGS. 7
and 8.
The method and the device for multichannel analog detection of the
invention can be used in a very general way for the detection of
numerous types of signals. However, they are particularly well
adapted to the carrying out of measurements of luminous flux with
photodiodes in the form of arrays or matrices.
It is known per se that in such a detector, each photodiode 1 is
equivalent to the mounting in parallel of a diode 2 and of a
capacitor 3. The current delivered to this circuit is a function of
the luminous flux received by the photodiode 1. A controllable
switch I.sub.1 makes it possible to read the charge of the
capacitor 3 at the chosen instant and thus to measure the luminous
energy received between two readings, that is to say the integral
of the flux.
The arrays or matrices of photodiodes incorporate a very large
number of elements of this type varying according to the
applications envisaged from 256 to several million elements. An
adapted device makes it possible to successively control switches
I.sub.1, I.sub.2, I.sub.3 . . . , I.sub.i . . . in such a way as to
make it possible to acquire signals which are representative of the
respective energies E.sub.1, E.sub.2, E.sub.3, . . . , E.sub.i, . .
. received by the corresponding photodiodes.
Generally, the signal to be detected is modulated in such a way as
to obtain a modulated signal S(P) of frequency f and of period T.
This signal is, needless to say, accompanied by noise.
This modulation frequency f can be relatively high, for example
several hundreds of MHz since, as will be seen in what follows, the
frequency of reading of the data delivered by the multipoint
receiver and digitizing of these data is greatly reduced with
respect to the frequency f.
In certain particular cases, the physical phenomenon observed
directly produces a periodic signal in the frequency range f
indicated above and can be directly exploited. Most often, the
signal to be detected is first of all deliberately modulated as for
the application according to the known principles of synchronous
detection for a single detector or of heterodyne detection. This
modulation can be obtained by numerous devices. When the signal to
be detected is a luminous signal carried by a beam produced by a
luminous source, it is possible to directly modulate the supply of
the source in order to obtain the modulated signal. It is equally
possible to use a means of modulation interposed between the source
and the multipoint receiver, for example a POCHELS module,
mechanical modulator, elasto-optical modulator, rotating polarizer,
etc.
The modulated signal S(P) represented for example in FIG. 2 is next
subjected to a synchronous attenuation of phase .PHI. represented
by A(.PHI.) in such a way as to produce an attenuated modulated
signal. This attenuation signal A(.PHI.) is a periodic signal of
the same frequency f as the modulated signal S(P) and it is
possible to vary its phase s determining its phase-shift with
respect to S(P). While fulfilling these conditions, it can assume
variable forms, can be constituted by a succession of square waves,
can itself be sinusoidal, etc.
The description which will follow, with reference to FIGS. 2 and 3,
relates to the case of a function A(.PHI.) of square waves.
This function A(.PHI.) has the value 1 during a fraction 1/n of the
period and a zero value otherwise. The attenuated modulated signal
S(P).times.A(.PHI.) thus corresponds to a chopping of a fraction
1/n of each period of the signal S(P) , with constant phase (on
FIG. 2, n=4).
The luminous signal thus modulated and attenuated is addressed to
the multipoint receiver which converts, for each of the points, the
luminous signal received into a primary analog electrical signal of
the same form.
This primary analog electrical signal is integrated over N periods
and produces a value V(P,.PHI.).
For each point P, the value V(P,.PHI.) is thus produced by the
luminous flux received by the detector at this point during N
periods of the signal S(P).
The values V(P,.PHI.) for the set of points, for a fixed phase
value .PHI., are then read, digitized and stored. These operations
of reading, digitizing, and storing are thus carried out at the
frequency f/N.
After having carried out these operations for a given .PHI. value,
this phase .PHI. of the synchronous attenuation A(.PHI.) is
increased with respect to the modulated signal S(P) by 2.pi./n.
A new chopping of the signal S(P) is then produced corresponding to
the time interval of 2.pi./n following the interval previously
chopped.
In the same way, the attenuated modulated signal is received by the
multipoint receiver which produces for each point a primary analog
electrical signal. This new primary analog signal is then also
integrated over N periods and produces for each point a value
V(P,.PHI.) corresponding to the new value of .PHI.. These new
values V(P,.PHI.) are again read, digitized and stored.
The phase .PHI. is successively increased by 2.pi./n thus storing n
sets of values V(P,.PHI.).
It is important to emphasize that in these operations, the reading
and the digitizing of the values V(P,.PHI.) have been produced at
the frequency f/N, thus at a frequency which can be much lower than
the frequency f. It is thus possible, as has been indicated above,
to dissociate the modulation frequency of the signal to be detected
f from the frequency of reading and of digitizing of the signals
delivered by the multipoint receiver. This feature makes it
possible to considerably improve the signal/noise ratio of the
detection and also makes it possible to employ a method of
heterodyne detection with any multipoint detector and in particular
with detectors of luminous flux of the CCD type.
When the multipoint receiver is composed of CCD elements, the
integration of the primary analog signal is obtained directly at
the level of each photodiode. The reading of the values V(P,.PHI.)
for each value of the phase shift .PHI. is obtained by the action
of a shift register 41 which controls the switches 40-1, 40-2, . .
. , 40-i, . . . and produces a signal on the video line 42. After
amplification by the amplifier 43, the video signal carrying the
informations V(P,.PHI.) is digitized by the analog-digital
converter 44 and stored in an ADP device 45.
A clock 46 ensures synchronization of the attenuation with the
modulated signal and controls the reading of the values
V(P,.PHI.).
Thus, a value V(P,.PHI.) has been stored for each of the points P.
In a way known per se, the processing unit 46 processes these n
values in such a way as to produce data which are representative of
the amplitude and of the phase of the signal S(P).
These values are then available for any purpose, they can be
displayed on a screen, represented with the aid of a plotter,
control any device, etc.
The above description has been given with reference to the use of
an attenuation represented by a square-wave function, any other
positive periodic function can be used to this effect so long as n
sets of values V(P,.PHI.) are produced for values .PHI. of the
phase shift of the attenuation with respect to the modulated signal
varying by 2.pi./n.
In FIG. 2, n=4. Higher values of n make it possible to measure
harmonics of the signal at a frequency which is a multiple of
f.
FIGS. 4 and 5 are two representations of a detection device
employing the invention. A luminous source 50 is used for the
observation of a physical phenomenon giving rise to a signal to be
detected. The luminous beam 51 produced by this source is detected
by a multipoint detector 52, constituted for example by a CCD
(charge coupled device) array of photodiodes. Between the source 50
and the multipoint receiver 52 are placed means of modulation 53 of
the luminous beam.
In the example of embodiment presented in FIG. 4, the attenuation
is produced by the shutter 54 while in the example represented in
FIG. 5, the attenuation is produced by acting on the supply 55 of
the luminous source 50. In the two embodiments, the means of
modulation 53 and the means of attenuation 54 or 55 are controlled
by the same clock 56 which ensures their synchronization. This same
clock 56 guides, by the agency of the driver module 57, the reading
of the array of diodes 52. The video signal carrying the values
V(P,.PHI.) for each value of the phase shift .PHI. is addressed to
the processing unit 58.
The application of the detection method of the invention for the
measurement of roughness by interferential microscopy will now be
described with reference to FIGS. 6, 7 and 8.
FIG. 7A is the general optical diagram, FIGS. 7B and 7C are
representations of the object 70 and of the multipoint detector in
their plane.
The object 70 whose roughness is measured is a reflector. The
source 71 produces, after crossing a WOLLASTON prism 72, two beams
73 and 74 polarized perpendicularly and focused by the effect of
the ocular 75, of the tube lens 76 and of the objective 77 onto the
object 70. A cylindrical lens 78 deforms the beams in such a way
that they focus along lines 79 and 80.
A separator cube 81 steers the reflected beams towards the
multipoint receiver 82 constituted by an array of CCD photodiodes.
A lens 83 brings about the focusing of the beams on the receiver,
an elasto-optic modulator 84 modulates the phase difference between
the two beams and a polarizer 85 produces their interference.
Thus, for each pair of points of the lines 79 and 80 and of the
object 70, the phase difference between the reflected beams depends
on the path difference introduced at the measurement points, that
is to say on their relative positions with respect to a plane which
is normal to the beam.
The signal received at the point 88 of the receiver 82
corresponding to the points 86, 87 (the points 86, 87 and 88 are
optically conjugated) depends on the phase difference between the
beams and thus on the relative positions of the points 86, 87 with
respect to the plane which is normal to the beam. The set of
signals received at the different points of the detector 82 makes
possible the measurement of the roughness of the object 70 along
lines 79, 80. By moving the object 70, perpendicularly to these
lines, the whole of its surface is scanned.
The microcomputer 100 drives the measurement, keeps the values
V(P,.PHI.) delivered by the multipoint detector 82 in the memory
and produces the measured values which are representative of
S(P).
The clock 101 controls and synchronizes, via the agency of the
block 102 for shaping and for processing the modulation signal, the
attenuation and the reading of the CCD register 103 associated with
the array of photodiodes 82.
The modulation is produced by the control of the modulator 84 at
the frequency of 50 KHz.
The attenuation is produced by the control of the supply 104 of the
source 105 by the block 102. The attenuation signal is composed of
two periodic signals offset by .pi. sent sequentially to the input
of the supply 104 (A(.PHI.) is a square-wave function, n=2).
The reading of the CCD register 103 is done for each phase of the
attenuation signal after storage of several exposures, for example
1000 (n=1000). The reading is thus done at 50 Hz.
The video signal produced by the reading of the CCD register 103 is
addressed to the analog-digital converter 106 and the values read
stored in the microcomputer 100.
The latter, producing the difference between the values obtained
during two readings for attenuations in phase opposition, extracts,
for each point of the detector, the amplitude of the signal.
FIG. 8 is an example of a display of the roughness of a sample
obtained with the device described.
The method and the device for multichannel detection described can
be applied in numerous fields. In particular, they can be used in
spectrometers, dichrometers, polarimeters, ellipsometers, optical
and scanning electron microscopes etc.
* * * * *